TW202242728A - Water quality management device, water quality management method, and water quality management program - Google Patents

Abstract

Provided are a water quality management device, a water quality management method, and a water quality management program with which it is possible to prevent the residual chlorine concentration in seawater from exceeding a reference value at a seawater discharge port. This water quality management device for use in a seawater utilization plant comprises: an attribute value acquisition unit that acquires an attribute value pertaining to seawater flowing through a condenser installed in the seawater utilization plant; a concentration prediction unit that, on the basis of the attribute value, predicts the residual chlorine concentration at a discharge port of a discharge path through which seawater is discharged from the condenser to the sea; a required amount calculation unit that, on the basis of the predicted residual chlorine concentration, calculates the required amount of a residual chlorine neutralizer to be injected into the discharge path; and a neutralizer injection unit that injects the neutralizer into the discharge path in the required amount.

Description

Water quality management device, water quality management method and water quality management program

The invention relates to a water quality management device, a water quality management method and a water quality management program. More specifically, it relates to a water quality management device, a water quality management method, and a water quality management program used in seawater utilization factories such as power plants.

As a countermeasure against accumulating organisms such as barnacles and mussels and biofilms attached to seawater systems of seawater utilization plants represented by thermal power plants and nuclear power plants, electrolytic chlorine (sodium hypochlorite) is generated and injected into water intakes widely. Technology.

For example, Patent Document 1 (Japanese Patent No. 4932529) discloses the following technology: sodium hypochlorite is generated by electrolysis of natural seawater, and the electrolyte solution containing the sodium hypochlorite is injected into the water intake of seawater to prevent the adhesion of marine organisms.

When injecting seawater electrolytic chlorine into the water intake, it is necessary to inject such that the residual chlorine concentration in the seawater discharge port does not exceed the reference value, but the residual chlorine concentration after the seawater electrolytic chlorine is injected depends on the water temperature and water quality. On the other hand, if you want to maintain the residual chlorine concentration that is effective for preventing fouling organisms, you may temporarily exceed the reference value at the discharge port. Therefore, it is desirable to prevent the residual chlorine concentration from exceeding the reference value by injecting an appropriate amount of neutralizing agent at an appropriate timing.

The present invention was made in view of the above problems, and an object of the present invention is to provide a water quality management device, a water quality management method, and a water quality management program capable of preventing the concentration of residual chlorine in seawater in a seawater discharge port from exceeding a reference value.

The present invention relates to a water quality management device used in a seawater utilization factory, comprising: an attribute value acquisition unit that acquires an attribute value of seawater circulating through a condenser set in the seawater utilization factory; and a concentration prediction unit based on the attribute value , predicting the concentration of residual chlorine in the water outlet of the water discharge channel that discharges the seawater from the condenser to the sea; the necessary amount calculation unit calculates the percentage of residual chlorine injected into the water discharge channel based on the predicted residual chlorine concentration a necessary amount of neutralizing agent; and, a neutralizing agent injecting part injecting the necessary amount of the neutralizing agent into the water discharge path.

In addition, preferably, the attribute value includes the residual chlorine concentration of the seawater in the water intake of the condenser, the water temperature of the seawater in the outlet of the condenser, and the distance from the water intake to the water outlet. The concentration predicting unit predicts the residual chlorine concentration in the water outlet by applying the attribute value to the Arrhenius equation (Arrhenius equation) for the time of seawater flowing down.

In addition, in the water quality management device, it is preferable that the concentration prediction unit includes an input data acquisition unit that acquires residual chlorine concentration, salinity, pH, ORP (oxidation Reduction potential), the historical record data of attribute values of water temperature as input data; the mark acquisition part acquires the historical record data of the residual chlorine concentration in the water outlet as a mark; the learning part uses the input data and the marked constructing a learning model for estimating the residual chlorine concentration in the water discharge port as learning data; The estimated value of the residual chlorine concentration mentioned above.

Furthermore, in the water quality management device, preferably, the learning unit constructs the learning model using a random forest.

In addition, in the water quality management device, it is preferable that the learning unit constructs the learning model using generalized addition (GAM method).

In addition, the present invention relates to a water quality management method used in a seawater utilization factory, comprising: an attribute value acquisition step of acquiring an attribute value of seawater circulating through a condenser set in the seawater utilization factory; a concentration prediction step based on the The attribute value is to predict the residual chlorine concentration in the water outlet of the water discharge channel that discharges the seawater from the condenser to the sea; the necessary amount calculation step is to calculate the residual chlorine injected into the water discharge channel based on the predicted residual chlorine concentration. a necessary amount of a neutralizing agent for chlorine; and a neutralizing agent injection step of injecting the necessary amount of the neutralizing agent into the water discharge path.

Furthermore, the present invention relates to a water quality management program used in a seawater utilization factory for causing a computer to execute: an attribute value acquisition step of acquiring an attribute value of seawater circulating through a condenser set in the seawater utilization factory; a concentration prediction step , based on the attribute value, predicting the residual chlorine concentration in the water outlet of the water discharge channel that releases the seawater from the condenser to the sea; the necessary amount calculation step is to calculate the concentration of chlorine injected into the A necessary amount of a neutralizing agent for residual chlorine in the water discharge path; and a neutralizer injecting step of injecting the necessary amount of the neutralizing agent into the water discharge path.

According to the present invention, it is possible to prevent the concentration of residual chlorine in the seawater in the discharge port of the seawater from exceeding the reference value.

Next, embodiments of the present invention will be described with reference to the drawings. [summary of the invention]

FIG. 1 is an overall configuration diagram of a seawater utilization factory 100 including a water quality management device 1 and a condenser 2 according to this embodiment mode. In the seawater utilization plant 100 , the condenser 2 takes out seawater as cooling water from the sea 3A through the intake channel 21 , and the cooling water after cooling the steam turbine etc. is discharged from the outlet 221 of the condenser 2 to the sea 3B through the discharge channel 22 . Inject the seawater electrolytic solution containing sodium hypochlorite at the chlorine injection point of the water intake channel 21 .

The water quality management device 1 predicts the residual chlorine concentration in the water discharge port 222 of the water discharge channel 22 based on the attribute value of the seawater flowing through the condenser 2, and calculates the amount of decrease compared with the residual chlorine concentration of the seawater in the condenser inlet 211, Thereafter, based on the reduced amount, the necessary amount of neutralizing agent is calculated, and the necessary amount of neutralizing agent is injected into the water discharge path 22 . [first embodiment]

Hereinafter, a water quality management device 1 as a first embodiment of the present invention will be described with reference to FIGS. 2 and 3 . [Structure of Embodiment]

FIG. 2 is a functional block diagram of the water quality management device 1 . The water quality management device 1 includes a control unit 11 , a neutralizer injection unit 12 , and a storage unit 13 .

The control unit 11 controls the entire water quality management device 1, and realizes various functions in this embodiment mode by appropriately reading and executing various programs from a storage area such as ROM, RAM, flash memory, or hard disk (HDD). The control unit 11 may be a CPU. The control unit 11 includes an attribute value acquisition unit 111 , a concentration prediction unit 112 , and a necessary amount calculation unit 113 .

In addition, the control unit 11 also includes general functional blocks such as a functional block for controlling the entire water quality management device 1 and a functional block for performing communication. However, since these general functional modules are well known to those skilled in the art, illustration and description are omitted.

The property value acquiring unit 111 acquires property values of seawater flowing through the condenser 2 . In more detail, for example, seawater is drawn from the water intake channel 21 through a water pump (not shown) set outside the water quality management device 1, and then passed through a water quality analysis device (not shown) set outside the water quality management device 1 shown) to analyze the drawn seawater. The attribute value acquisition part 111 acquires the attribute value of the water quality of seawater analyzed by the water quality analyzer. In addition, the attribute value acquiring unit 111 acquires the water temperature of the seawater in the outlet 221 of the condenser 2 , the time for the seawater to flow down from the condenser inlet 211 to the water outlet 222 , and the like as attribute values.

Here, as attribute values of water quality, for example, in addition to the residual chlorine concentration and water temperature of seawater in the condenser inlet 211, for the purpose of improving accuracy, organic matter concentration, the amount of salt contained in seawater, pH, ORP (oxidation reduction potential).

The concentration predicting unit 112 predicts the concentration of residual chlorine in the water outlet 222 of seawater based on the attribute value acquired by the attribute value acquiring unit 111 . In particular, in this embodiment, the concentration predicting unit 112 uses the residual chlorine concentration of seawater in the inlet 211 of the condenser 2, the water temperature of seawater in the outlet 221 of the condenser 2, and the concentration of chlorine from the condenser 2 as attribute values. The flow time of the seawater from the inlet 211 to the water outlet 222 is applied to the Arrheinius equation to estimate the residual chlorine concentration in the seawater water outlet 222 .

Here, the "Arenis equation" is a formula for predicting the rate of a chemical reaction at a certain temperature, and the reaction rate constant (k) expressing the reaction rate, as shown in the following formula 1, indicates that the temperature (T) is high, the activation energy (Ea ) formula that becomes larger when low.

Among them, A is a temperature-independent constant (pre-exponential factor), Ea is the activation energy per 1 mol, R is the gas constant, and T is the absolute temperature. Here, the "pre-exponential factor" refers to a factor indicating the number of collisions between molecules in a bimolecular reaction.

The necessary amount calculating unit 113 calculates the required amount of neutralizing agent for residual chlorine injected into the water discharge channel 22 based on the residual chlorine concentration predicted by the concentration predicting unit 112 . More specifically, the necessary amount calculation unit 113 calculates the necessary amount of the neutralizing agent based on the amount of decrease in the estimated residual chlorine concentration in the water discharge port 222 compared to the residual chlorine concentration in the water intake port 211 .

Here, the neutralizing agent may be, for example, 35% hydrogen peroxide water, sodium sulfite or sodium thiosulfate, etc., which can quickly neutralize residual chlorine. In addition, the reaction by-products when using 35% hydrogen peroxide water are oxygen, water, and chloride ions, and the reaction by-products when using sodium sulfite are sulfate ions and chloride ions. These are found in abundance in seawater.

The neutralizing agent injecting unit 12 injects the necessary amount of neutralizing agent calculated by the required amount calculating unit 113 into the water discharge channel 22 . In particular, it is preferable to automatically control the injection of the neutralizing agent into the water discharge channel 22 through the neutralizing agent injection part 12 .

The storage unit 13 stores the attribute value acquired by the attribute value acquiring unit 111 , the residual chlorine concentration predicted by the concentration predicting unit 112 , and the necessary amount of neutralizing agent calculated by the necessary amount calculating unit 113 . [Action of the embodiment]

FIG. 3 is a flowchart showing the operation of the water quality management device 1 .

In step S1, the attribute value acquisition unit 111 acquires the attribute value of the water quality in the intake port 211 of the seawater, the water temperature of the seawater in the outlet 221 of the condenser 2, and the flow time of the seawater from the inlet 211 to the discharge port 222, etc. .

In step S2 , the concentration predicting unit 112 predicts the concentration of residual chlorine in the water outlet 222 of seawater based on the attribute value acquired by the attribute value acquiring unit 111 .

In step S3 , the necessary amount calculating unit 113 calculates the required amount of neutralizing agent for residual chlorine injected into the water discharge channel 22 based on the residual chlorine concentration predicted by the concentration predicting unit 112 .

In step S4 , the neutralizing agent injecting unit 12 injects the necessary amount of neutralizing agent calculated by the required amount calculating unit 113 into the water discharge path 22 . [Effect of the embodiment]

The water quality management device 1 of the present embodiment is a water quality management device 1 used in a seawater utilization factory 100, and includes: an attribute value acquisition unit 111, which acquires an attribute value of seawater circulating in the condenser 2 set in the seawater utilization factory 100; The predicting part 112 predicts the residual chlorine concentration in the water discharge port 222 of the water discharge channel 22 that discharges seawater from the condenser 2 to the sea based on the attribute value; 22, the necessary amount of neutralizing agent for residual chlorine;

Thereby, it can prevent that the residual chlorine concentration in the seawater in the discharge port of seawater exceeds a reference value. In particular, before the residual chlorine concentration in the water discharge port 222 exceeds the reference value, it can be prevented from exceeding the reference value by injecting an appropriate amount of neutralizing agent. In this way, the base value of the electrolytic chlorine injection concentration can be adjusted to a higher level, and the heat exchange efficiency of the condenser 2 can be improved by more effectively suppressing the adhesion of fouling organisms and biofilms that hinder power generation, and a great cost reduction effect can be obtained.

In addition, in the water quality management device 1, the above attribute values include the residual chlorine concentration of the seawater in the inlet 211 of the condenser 2, the water temperature of the seawater in the outlet 221 of the condenser 2, and the seawater from the outlet 221 to the water outlet 222. The concentration predicting unit predicts the residual chlorine concentration in the water discharge port 222 by applying the above attribute values to the Arenius equation.

Thus, by using the Arenius equation, it is possible to prevent the residual chlorine concentration in seawater from exceeding the reference value. [Second embodiment]

Hereinafter, a water quality management device 1A as a second embodiment of the present invention will be described with reference to FIGS. 4 to 6 . In addition, below, for simplification of description, the difference between 1A of water quality management apparatuses and the water quality management apparatus 1 is mainly demonstrated. [Composition of Execution Power]

The basic configuration of the water quality management device 1A is the same as that of the water quality management device 1 shown in FIG. 2 . However, the water quality management device 1A includes a concentration prediction unit 112A instead of the concentration prediction unit 112 included in the water quality management device 1 . The concentration prediction unit 112 mainly predicts the residual chlorine concentration in the water discharge port 222 by using the Arenius equation, but the concentration prediction unit 112A predicts the residual chlorine concentration in the water discharge port 222 through machine learning.

FIG. 4 is a functional block diagram of the density prediction unit 112A. The concentration prediction unit 112A includes an input data acquisition unit 114 , a flag acquisition unit 115 , a learning unit 116 , and an estimated value generation unit 117 .

The input data acquisition unit 114 acquires historical record data including attribute values of residual chlorine concentration, salinity, pH, ORP (oxidation-reduction potential), and water temperature of seawater in the water intake 211 from the storage unit 13 as input for machine learning. data.

The marker acquisition unit 115 acquires the history data of the residual chlorine concentration in the water outlet 222 from the storage unit 13 as a marker for machine learning.

The learning unit 116 constructs a learning model for estimating the residual chlorine concentration in the water discharge port 222 by performing machine learning using a set of input data and labels as learning data, and stores the constructed learning model in the storage unit 13 .

Here, when a set of input data and a label is used as learning data, the time for seawater to flow down from the outlet 221 of the condenser 2 to the water outlet 222 is considered, and the data at the inlet of the condenser is paired with the water outlet data after this time. mode for processing. In addition, for example, both the target variable and the explanatory variable can treat negative values as outliers and remove them.

In addition, the machine learning performed by the learning unit 116 may be random forest or generalized addition (GAM method). Here, "Random Forest" is one of the algorithms of machine learning. It is an algorithm that uses decision trees as weak learners and synthesizes multiple weak learners learned from randomly sampled training data to improve generalization ability. In addition, "generalized addition" refers to an algorithm using a model in which linear predictors in a generalized linear model are the sum of nonlinear functions, and local regression functions, smoothing splines, B-splines, natural splines etc. Among them, an algorithm using a smoothing spline as a nonlinear function is called "GAM method".

The estimated value generation unit 117 constructs a learning model through the learning unit 116, and after the constructed learning model is stored in the storage unit 13, acquires the learning model from the storage unit 13, and applies new attribute values to the learning model, thereby generating The estimated value of the residual chlorine concentration. [Action of the embodiment]

FIG. 5 is a flowchart showing operations during machine learning of the water quality management device 1A.

In step S11 , the input data acquisition unit 114 acquires history data including attribute values of seawater residual chlorine concentration, salinity, pH, ORP (oxidation-reduction potential), water temperature, and flow rate from the storage unit 13 as input data.

In step S12, the flag acquisition part 115 acquires the history data of the residual chlorine concentration in a water outlet as a flag.

In step S13 , the learning unit 116 uses a set of input data and labels as learning data.

In step S14, the learning unit 116 performs machine learning using the learning data.

In step S15, when machine learning is complete|finished (S15: YES), a process transfers to step S16. When the machine learning has not been completed ( S15 : NO), the process proceeds to step S11 .

In step S16 , the learning unit 116 stores the constructed learning model in the storage unit 13 .

FIG. 6 is a flowchart showing the operation of the water quality management device 1A at the time of injection of the neutralizing agent.

In step S21 , the estimated value generation unit 117 acquires the learning model from the storage unit 13 .

In step S22 , the estimated value generator 117 acquires a new attribute value from the attribute value acquirer 111 .

In step S23 , the estimated value generator 117 generates an estimated value (predicted value) of the residual chlorine concentration in the water outlet 222 by applying the new attribute value to the learning model.

In step S24 , the necessary amount calculating unit 113 calculates the required amount of neutralizing agent for residual chlorine injected into the water discharge channel 22 based on the residual chlorine concentration predicted by the concentration predicting unit 112 .

In step S25 , the neutralizing agent injecting unit 12 injects the necessary amount of neutralizing agent calculated by the required amount calculating unit 113 into the water discharge path 22 . [Effect of the embodiment]

In the water quality management device 1A, the concentration prediction unit 112A includes an input data acquisition unit 114 for acquiring attribute values including residual chlorine concentration, salt content, pH, ORP (oxidation-reduction potential), and water temperature of seawater in the water intake 211. The historical record data is as input data; Mark acquisition part 115, obtains the historical record data of the residual chlorine concentration in the water discharge port 222 as mark; The learning model of the residual chlorine concentration; and the estimated value generating unit 117 generates an estimated value of the residual chlorine concentration by applying new attribute values to the learning model after the learning model is constructed.

Thereby, it is possible to inject the neutralizing agent into the water discharge channel 22 based on a more accurate prediction value of the residual chlorine concentration.

In addition, the learning unit 116 forms a learning model through a random forest.

As a result, correspondence with a plurality of explanatory variables and high-speed learning can be performed, and the degree of importance (degree of contribution) of the explanatory variables can be calculated.

Also, the learning section 116 constructs a learning model using generalized addition.

In this way, by using a complex form of function, it is possible to explain a case that is not a simple proportional relationship, and it is possible to improve the accuracy of prediction while maintaining the explanatory properties of a linear model. [forecast data]

[Arenis Equation] (1) Analysis of data from June 29, 2018 to November 9, 2018 Using the measurement data of A power plant No. 1 unit (maximum output 340MW) from June 29, 2018 to November 9, 2018

(1-1) Relationship with the concentration at the inlet of the condenser

It is also known that the decay of the residual chlorine concentration contributes greatly to the initial concentration. For the Arenis equation shown in (1-1), the residual chlorine concentration at the inlet of the condenser is divided into 0.05mg/L or more, 0.03mg/L or more and less than 0.05mg/L, and less than 0.03mg/L according to the output power of the power generation. For the three cases of L, Fig. 7A to Fig. 9C show the Arenis equations in various cases.

The higher the residual chlorine concentration, the higher the coefficient of determination at any power generation output. The highest coefficient of determination is 0.589 when the power generation output is above 200MW and the residual chlorine concentration is above 0.05mg/L. In addition, in the case where the residual chlorine concentration is less than 0.03 mg/L, the coefficient of determination is lower than 0.05 in any case of power generation output.

[machine learning] As a method for predicting the residual chlorine concentration in the water outlet from the residual chlorine concentration at the condenser inlet, a study was conducted on the possibility of application of machine learning.

(1) Data used The data used for model construction and forecasting uses the 1-minute value from June 29, 2018 to March 31, 2019 in Unit 1 of A Power Plant. The variables used and the processing method of the data are shown in Table 1 below.

[Table 1] target constant Drain residual chlorine concentration Explanatory variables condenser inlet Residual chlorine concentration, salinity, pH, ORP, water temperature, electrolysis output power Condenser outlet water temperature A, B Data processing method 1. The data of the inlet of the condenser is paired with the data of the water outlet after 4 minutes for processing. 2. Negative values of the target variable and explanatory variable are regarded as outliers and removed. 3. The concentration of residual chlorine in the condenser is removed based on a Faraday cell The data after 5 minutes of cleaning. 4. For the data whose residual chlorine concentration after removal is 0.1 or more, the data whose difference with the moving average value 30 minutes after washing is positive is excluded. Data 6. Values exceeding 1.5 times the interquartile range of the variables used are considered outliers and removed

(2) The results of the prediction model Using the processed data, predictions were made using two prediction models of random forest and generalized additive model (GAM).

(2-1) Prediction results of random forest Using the data from June 29, 2018 to February 28, 2019 as the learning data, predictions were made for March 1, 2019 to March 31, 2019. Figures 10A and 10B show a comparison of raw data and predicted results.

Raw data and predicted results show the same move. Relative to the original data, the convergence is roughly in the range of 0.01mg/L, but due to the deviation in the distribution of errors, the coefficient of determination between the original data and the predicted results is as low as 0.096.

(2-2) Prediction results of generalized addition (GAM) Three forecasts were made by varying the learning horizon and forecast horizon.

[Prediction result 1] Similar to the prediction using random forest, the data from June 29, 2018 to February 28, 2019 were used as learning data, and the prediction was made for March 1, 2019 to March 31, 2019 . 11A and 11B show a comparison of raw data and predicted results.

Same as random forest results, raw data and predicted results show the same behavior. Compared with the original data, it is roughly converged in the range below 0.01mg/L, and the deviation of the error distribution is also small. Therefore, the coefficient of determination between the original data and the predicted result is 0.272, which is higher than the result of the random forest.

[Prediction result 2] Change the learning period and prediction period differently from the above-mentioned prediction, and use the data from June 29, 2018 to October 31, 2018 and December 1, 2018 to March 31, 2019 as the learning data , made a forecast from November 1, 2018 to November 30, 2018. 12A and 12B show a comparison of raw data and predicted results.

Same as Prediction 1, raw data and prediction show the same behavior. Relative to the original data, the convergence is roughly in the range of 0.01mg/L, and the coefficient of determination between the original data and the predicted results is 0.303.

[Prediction result 3] Using the data from February 1, 2019 to February 14, 2019 as the learning data, the prediction for February 15, 2019 to February 20, 2019 was performed. 13A and 13B show a comparison of raw data and predicted results.

Raw data behaved roughly in line with predicted results, showing high reproducibility. Compared with the original data, the coefficient of determination between the original data and the predicted result is 0.505, which is quite high. [Summarize]

When the power generation output is 200 MW or more and the residual chlorine concentration at the condenser inlet is 0.05 mg/L or more, it can be confirmed that the coefficient of determination of the approximate curve based on the Arenis equation is the highest for the decay reaction from the condenser inlet to the water outlet , the lower the power generation output and the residual chlorine concentration are, the lower the coefficient of determination tends to be. Therefore, it can be considered that the higher the output power of the power generation and the higher the concentration at the inlet of the condenser, the more effective the prediction of the concentration at the water outlet based on the Arrhenius formula is.

Regarding the prediction of water outlet concentration based on machine learning, as a result of verification through multiple models and conditions, it was confirmed that the prediction accuracy is almost practical, and it can be applied.

The management method of the water quality management device 1 or 1A is implemented through software. When realizing by software, the program which comprises this software is installed in a computer (water quality management apparatus 1 or 1A). In addition, these programs can be distributed to users by recording them on removable media, or by downloading them to users' computers via the Internet. In addition, these programs may be provided to the user's computer (the water quality management device 1 or 1A) as a Web service via a network without downloading.

1. 1A: Water quality management device 2: Condenser 100:Seawater Utilization Factory 11: Control Department 111: attribute value acquisition part 112, 112A: Concentration Prediction Department 113:Necessary Quantity Calculation Department 114: input data acquisition part 115: Mark Acquisition Department 116: Learning Department 117: Estimated value generation unit 12: Neutralizer injection part 13: Storage department 21: water intake 211: Condenser inlet 22: Release the waterway 221: Export 222: water outlet 3A, 3B: Sea S1, S2, S3, S4: steps S11, S12, S13, S14, S15, S16: steps S21, S22, S23, S24, S25: steps

FIG. 1 is an overall structure diagram of a seawater utilization plant according to an embodiment of the present invention. Fig. 2 is a functional block diagram of a water quality management device according to an embodiment of the present invention. Fig. 3 is a flowchart showing the operation of the water quality management device according to one embodiment of the present invention. Fig. 4 is a functional block diagram of a concentration prediction unit included in the water quality management device according to the embodiment of the present invention. Fig. 5 is a flowchart showing the operation of the water quality management device according to one embodiment of the present invention. Fig. 6 is a flowchart showing the operation of the water quality management device according to one embodiment of the present invention. FIG. 7A is a graph showing the Arrhenius equation of an embodiment of the present invention. FIG. 7B is a graph showing the Arrhenius equation of an embodiment of the present invention. FIG. 7C is a graph showing the Arenius equation of an embodiment of the present invention. FIG. 8A is a graph showing the Arenius equation of an embodiment of the present invention. FIG. 8B is a graph showing the Arrhenius equation of an embodiment of the present invention. FIG. 8C is a graph showing the Arrhenius equation of an embodiment of the present invention. FIG. 9A is a graph showing the Arenius equation of an embodiment of the present invention. FIG. 9B is a graph showing the Arenius equation of an embodiment of the present invention. FIG. 9C is a graph showing the Arrhenius equation of an embodiment of the present invention. FIG. 10A is a graph showing a comparison of raw data and predicted results based on random forest according to an embodiment of the present invention. FIG. 10B is a graph showing a comparison of raw data and predicted results based on random forest according to an embodiment of the present invention. FIG. 11A is a graph showing a comparison of raw data and predicted results based on generalized addition (GAM) according to an embodiment of the present invention. FIG. 11B is a graph showing a comparison of raw data and predicted results based on generalized addition (GAM) according to an embodiment of the present invention. FIG. 12A is a graph showing a comparison of raw data and predicted results based on generalized addition (GAM) according to an embodiment of the present invention. FIG. 12B is a graph showing a comparison of raw data and predicted results based on generalized addition (GAM) according to an embodiment of the present invention. FIG. 13A is a graph showing a comparison of raw data and predicted results based on generalized addition (GAM) according to an embodiment of the present invention. FIG. 13B is a graph showing a comparison of raw data and predicted results based on generalized addition (GAM) according to an embodiment of the present invention.

1: Water quality management device

11: Control Department

111: attribute value acquisition part

112: Concentration Prediction Department

113:Necessary Quantity Calculation Department

12: Neutralizer injection part

13: Storage department

Claims (7)

A water quality management device for a seawater utilization factory, comprising: an attribute value acquiring unit that acquires an attribute value of seawater circulating through a condenser set in the seawater utilization factory; a concentration predicting unit predicting a residual chlorine concentration in a discharge port of a discharge channel that discharges the seawater from the condenser to the sea based on the attribute value; a necessary amount calculation section that calculates a necessary amount of a neutralizing agent for residual chlorine injected into the water discharge path based on the predicted residual chlorine concentration; and A neutralizing agent injecting part injects the necessary amount of the neutralizing agent into the water discharge path. The water quality management device according to claim 1, wherein the attribute value includes a residual chlorine concentration of seawater in a water intake of the condenser, a water temperature of seawater in an outlet of the condenser, and The concentration predicting unit predicts the residual chlorine concentration in the water outlet by applying the attribute value to the Arrhenius equation for the time from the water intake to the water outlet. The water quality management device according to claim 1, wherein the concentration prediction unit includes: an input data acquisition unit that acquires historical record data including a residual chlorine concentration, salinity, pH, ORP (oxidation-reduction potential), and water temperature attribute values of seawater in the inlet of the condenser as an input data; a mark acquisition part, which acquires the historical record data of the residual chlorine concentration in the water outlet as a mark; a learning unit that uses the set of the input data and the flag as learning data to construct a learning model for estimating the residual chlorine concentration in the water outlet; and An estimated value generating unit generates an estimated value of the residual chlorine concentration by applying new attribute values to the learning model after the learning model is constructed. The water quality management device according to claim 3, wherein the learning unit constructs the learning model using a random forest. The water quality management device according to claim 3, wherein the learning unit constructs the learning model using generalized addition. A water quality management method for a seawater utilization factory, comprising: an attribute value acquiring step of acquiring an attribute value of seawater circulating in a condenser set in the seawater utilization factory; A concentration predicting step, based on the attribute value, predicting a residual chlorine concentration in a discharge port of a discharge channel that discharges the seawater from the condenser to the sea; a necessary amount calculation step of calculating a necessary amount of a neutralizing agent for residual chlorine injected into said water discharge path based on the predicted residual chlorine concentration; and A neutralizing agent injection step, injecting the necessary amount of the neutralizing agent into the water discharge channel. A water quality management program for a seawater utilization factory for causing a computer to execute: an attribute value acquiring step of acquiring an attribute value of seawater circulating in a condenser set in the seawater utilization factory; A concentration predicting step, based on the attribute value, predicting a residual chlorine concentration in a discharge port of a discharge channel that discharges the seawater from the condenser to the sea; a necessary amount calculation step of calculating a necessary amount of a neutralizing agent for residual chlorine injected into said water discharge path based on the predicted residual chlorine concentration; and A neutralizing agent injection step, injecting the necessary amount of the neutralizing agent into the water discharge channel.

Images